Absolute quantification of cerebral tissue oxygen saturation with multidistance broadband NIRS in newborn brain

Tissue oximetry with near-infrared spectroscopy (NIRS) is a technique for the measurement of absolute tissue oxygen saturation (StO(2)). Offering a real-time and non-invasive assessment of brain oxygenation and haemodynamics, StO(2) has potential to be used for the assessment of newborn brain injury. Multiple algorithms have been developed to measure StO(2), however, issues with low measurement accuracy or extracranial tissue signal contamination remain. In this work, we present a novel algorithm to recover StO(2) in the neonate, broadband multidistance oximetry (BRUNO), based on a measurement of the gradient of attenuation against distance measured with broadband NIRS. The performance of the algorithm was compared to two other published algorithms, broadband fitting (BF) and spatially resolved spectroscopy (SRS). The median error when recovering StO(2) in light transport simulations on a neonatal head mesh was 0.4% with BRUNO, 4.2% with BF and 9.5% with SRS. BRUNO was more sensitive to brain tissue oxygenation changes, shown in layered head model simulations. Comparison of algorithm performance during full oxygenation-deoxygenation cycles in a homogeneous dynamic blood phantom showed significant differences in the dynamic range of the algorithms; BRUNO recovered StO(2) over 0–100%, BF over 0–90% and SRS over 39–80%. Recovering StO(2) from data collected in a neonate treated at the neonatal intensive care showed different baseline values; mean StO(2) was 64.9% with BRUNO, 67.2% with BF and 73.2% with SRS. These findings highlight the effect of StO(2) algorithm selection on oxygenation recovery; applying BRUNO in the clinical care setting could reveal further insight into complex haemodynamic processes occurring during neonatal brain injury.